Soluble forms of the HIV-1 receptor CD4 have been studied as potential therapeutics for a number of years. Soluble CD4 (sCD4), and the more bioavailable fusion protein, CD4-Ig, neutralize more broadly than any neutralizing antibody, and, unlike antibodies, they can irreversibly inactivate the envelope glycoprotein trimer. Nonetheless, several problems with CD4-Ig limit its therapeutic use. First, the affinity of CD4 for the HIV-1 envelope glycoprotein is lower than that of efficient neutralizing antibodies. Second, CD4-Ig, like sCD4, enhances HIV-1 infection at low concentrations, a serious difficulty as local concentrations vary in vivo. This enhancement occurs because CD4-Ig bound at low valency promotes association of the envelope glycoprotein with the cellular coreceptor, CCR5 or CXCR4. Third, low concentration enhancement permits sufficient residual viral replication in the presence of CD4-Ig to enable viral escape. Finally, there remains the practical problem of using a protein therapeutic like CD4-Ig in resource-poor settings. The first goal of this proposal is to further develop a more effective form of CD4-Ig. Our key observation is that by adding a short peptide mimetic of CCR5, we substantially increase the affinity of CD4-Ig for the envelope glycoprotein while preventing the enhancement of infection observed with unmodified CD4-Ig. We hypothesize that this enhanced CD4-Ig (eCD4-Ig) will impose a significantly higher fitness cost on escaping virus than does unmodified CD4-Ig or neutralizing antibodies. Our second goal is to build on recent advances in gene delivery using self-complementary adeno-associated virus (scAAV) vectors. Increasingly these vectors are recognized as a safe, unobtrusive way to maintain high levels of protein therapeutics in vivo. For example, a single intramuscular injection with scAAV-vector expressing an scFv immunoadhesin or CD4-Ig can sustain high serum protein levels for more than two years and protect a macaque from an SIV challenge. We hypothesize that, if key hurdles can be overcome, scAAV-delivered protein inhibitors like eCD4-Ig could be used to drive down viral loads to the point where transmission is unlikely. However, these hurdles are substantial. First, immune clearance of scAAV-delivered scFv-Ig and CD4-Ig is a frequent and formidable problem in healthy, uninfected macaques, although perhaps less of a concern in late-stage infections. A second major problem is that of viral escape, a much greater concern with therapeutic applications of scAAV than with prophylactic ones. We will address both of these issues as we compare the efficacy of scAAV-delivered CD4-Ig and eCD4-Ig in a macaque model of infection. These studies will therefore (1) evaluate the use scAAV-delivered protein inhibitors in infected individuals while (2) developing a novel therapeutic with important advantages over neutralizing antibodies and unmodified CD4-Ig.